Chronic Nerve Growth Factor Exposure Increases Apoptosis in a Model of In Vitro Induced Conjunctival Myofibroblasts

In the conjunctiva, repeated or prolonged exposure to injury leads to tissue remodeling and fibrosis associated with dryness, lost of corneal transparency and defect of ocular function. At the site of injury, fibroblasts (FB) migrate and differentiate into myofibroblasts (myoFB), contributing to the healing process together with other cell types, cytokines and growth factors. While the physiological deletion of MyoFB is necessary to successfully end the healing process, myoFB prolonged survival characterizes the pathological process of fibrosis. The reason for myoFB persistence is poorly understood. Nerve Growth Factor (NGF), often increased in inflamed stromal conjunctiva, may represent an important molecule both in many inflammatory processes characterized by tissue remodeling and in promoting wound-healing and well-balanced repair in humans. NGF effects are mediated by the specific expression of the NGF neurotrophic tyrosine kinase receptor type 1 (trkANGFR) and/or the pan-neurotrophin glycoprotein receptor (p75NTR). Therefore, a conjunctival myoFB model (TGFβ1-induced myoFB) was developed and characterized for cell viability/proliferation as well as αSMA, p75NTR and trkANGFR expression. MyoFB were exposed to acute and chronic NGF treatment and examined for their p75NTR/trkANGFR, αSMA/TGFβ1 expression, and apoptosis. Both NGF treatments significantly increased the expression of p75NTR, associated with a deregulation of both αSMA/TGFβ1 genes. Acute and chronic NGF exposures induced apoptosis in p75NTR expressing myoFB, an effect counteracted by the specific trkANGFR and/or p75NTR inhibitors. Focused single p75NTR and double trkANGFR/p75NTR knocking-down experiments highlighted the role of p75NTR in NGF-induced apoptosis. Our current data indicate that NGF is able to trigger in vitro myoFB apoptosis, mainly via p75NTR. The trkANGFR/p75NTR ratio in favor of p75NTR characterizes this process. Due to the lack of effective pharmacological agents for balanced tissue repairs, these new findings suggest that NGF might be a suitable therapeutic tool in conditions with impaired tissue healing.


Introduction
Acute and chronic inflammation of the conjunctiva causes the alteration of local architecture with tissue remodeling and fibrosis associated with ocular dryness and corneal complications leading to visual function impairment [1]. Traumatic, chemical, inflammatory or infectious insults, as well as surgical scarring, all represent promoting-causes of ocular fibrosis [2]. Pharmacological management and/or surgical strategies are required to restore healthy conjunctival structure, function and ocular transparency of compromised cornea. Conjunctival Fibroblasts (FB) and their differentiated a Smooth Muscle Actin (aSMA) expressing myofibroblasts (myoFB) play a pivotal role during repair/remodeling processes as targets, effectors and modulators of the process [3]. MyoFB appearance at the early phases and disappearance at the late phases of tissue repair are necessary steps to gain proper tissue healing [4]. MyoFB persistence seems to be one of the causes of fibrosis and may be due to an ''acquired resistance to cell apoptosis'' and/or to a microenvironment suitable for FB prolonged activity and survival [5]. Apoptosis of myoFB has been documented in vivo, representing a physiological process [3]. More than a few soluble factors (TGFb1, IL6, IL4, IL13, etc.) have been reported to promote or counteract (IL1b, IL7) FB/myoFB physiological balance [6].
Several findings suggest the contribution of either endogenously produced or topically applied Nerve Growth Factor (NGF) in healing processes, wound-narrowing, tissue remodeling and fibrosis processes [6][7][8][9]. Tears, as well as inflamed conjunctiva, are characterized by increased TGFb1 and NGF, a consequence of the inflamed microenvironment [10,11]. NGF is a pleiotrophic factor that promotes cell growth, differentiation, survival and death among different tissues [7,8,11]. NGF activities appear to be mediated by two different receptors: the specific NGF neurotrophic tyrosine kinase receptor type 1 (trkA NGFR ) of 140 kDa and the pan-neurotrophin low affinity glycoprotein receptor (p75 NTR ) of 75 kDa, a typical death receptor belonging to the tumor necrosis receptor superfamily [8,12]. Our group previously reported that human conjunctival FB express constitutive trkA NGFR and differentiate into aSMA and p75 NTR bearing myoFB upon NGF stimulation, suggesting NGF contribution at wound-narrowing and during healing of ulcers [13,14]. Even though NGF withdrawal is largely considered the cause of NGFassociated cell death, several findings point to NGF as both a proand anti-apoptotic factor. A specific cell-surface trkA NGFR / p75 NTR ratio seems to be directly responsible for either proliferative and/or survival effects (trkA NGFR ) or apoptotic response (p75 NTR ), with p75 NTR acting alone or in combination to modulate trkA NGFR trafficking and/or signaling [12,15,16]. NGF-induced myoFB conversion was associated with the selective expression of p75 NTR [12].
To test whether NGF exposure might modulate myoFB behavior, we induced in vitro the myoFB phenotype via conjunctival FB exposure to TGFb1. Here we describe evidence that conjunctival TGFb1-induced myoFB express p75 NTR and are more sensitive to apoptosis after NGF treatment.

Acute and Chronic NGF Exposure Increases p75 NTR Expression but does not Influence trkA NGFR Expression
Conjunctival FB were exposed to 2 ng/mL TGFb1 for 3 days to develop the myoFB phenotype, according to previous studies [17]. Induced myoFB were re-plated at high density (to retain myoFB phenotype) and once at confluence were exposed to single (mimicking acute treatment) or three repeated NGF doses every two days (mimicking chronic treatment), with sampling at the specified times from the last stimulation. TGFb1-induced myoFB (herein referred to as myoFB) showed a significant high expression of aSMA protein, as detected by confocal microscopy, cell surface ELISA and flow cytometry ( Figure 1A). These myoFB expressed both trkA NGFR and p75 NTR , as shown by flow cytometry (Figure 1B-C; black lines).
When the myoFB were exposed to NGF (100 ng/mL) and harvested 24 hrs from the last stimulation, an increase of p75 NTR protein (Mean Fluorescent Intensity, MFI) was detected in both acute (DMFI = 90.25, range 57.00 to 173.20; Figure 1B) and more significant chronic (DMFI = 143.50, range 120.30 to 184.00; Figure 1C) treatments, as compared to untreated ones (DMFI = 58.00, range 56.00 to 60.80; p,.05; black lines). The morphological distribution of both trkA NGFR and p75 NTR in untreated, acute and chronic NGF-treated myoFB (48 hrs from last stimulus) is shown in Figure 1D. The molecular analysis carried out on myoFB exposed to increasing NGF (0-200 ng/mL) showed the upregulation of p75 NTR target gene as soon as 5 hrs from stimulation, while unchanged values characterized trkA NGFR expression ( Figure 1E). p75 NTR expression showed a linear increase between 1 and 100 ng/mL (R 2 = 0.996, p75 NTR ratios vs. NGF doses). The maximum p75 NTR increase was detected at 100 ng/mL NGF (trkA NGFR : 2.7160. 12  p75 NTR over-expression was associated with aSMA protein (data not shown) and mRNA down-regulation, especially after chronic NGF exposure ( Figure 1F). A significant 6.8-fold decrease in aSMA expression was detected after 100 ng/mL chronic NGF exposure while a slight decrease was observed at all the other concentrations (1 ng/mL NGF: 1.30-fold decrease, 10 ng/mL NGF: 1.10-fold decrease, 200 ng/mL NGF: 2.00-fold decrease; p,.05). aSMA-expressing myoFB showed p75 NTR mainly localized at the nuclear membrane or along the cytoskeleton (data not shown). A direct correlation was found between aSMA and p75 NTR upon chronic NGF exposure (Rho = 0.702; p = .002, Spearman's rank test), implying that the decrease of both markers occurred alongside chronic NGF exposure.

Acute and Chronic NGF Exposure Triggers TUNEL and AnnexinV Positive Cells
Since a decrease in the number of cells and viability were detected after acute and chronic NGF exposure, NGF exposed monolayers were stained with HO342 and/or DAPI and observed by fluorescence microscopy. Cells with clear signs of nuclear fragmentation were visible as soon as 5 hrs after acute NGF exposure. As shown, acute ( Figure 2A) and chronic ( Figure 2B) NGF exposure increased the number of rounded, crescent and condensed cells as well as the blabbed chromatine, indicating that myoFB were undergoing apoptosis in a time and a dose dependent manner.
TUNEL-reactivity confirmed the presence of in situ oligonucleosomal fragmentation. TUNEL-positive cells were quantified after acute and chronic NGF treatment, by counting in a blind fashion those myoFB showing brown-picnotic nuclei in different random fields ( Figure 2C). Data were statistically compared to those of untreated myoFB ( Figure 2D). Since some apoptotic nuclei were devoid of visible cytoplasm, the number of apoptotic cells was likely to be underestimated ( Figure 2D vs. Figure 2A,B). Few TUNEL-positive myoFB were also detected in untreated cultures of myoFB (data not shown).
Additional studies were performed by using AnnexinV/caspase staining. Outer membrane exposure of phosphatidylserine (PS) was detected by incubating unfixed cells with FC-conjugated AnnexinV. As shown, AnnexinV positivity increased in myoFB after acute ( Figure 3A) and chronic ( Figure 3B) NGF treatments. The intensity of FC-AnnexinV binding was quantified by flow cytometry (MFI), resulting in a significant increase of % MFI after acute NGF exposure and more consistent after chronic NGF exposure, as compared to untreated cells ( Figure 3C). As shown by confocal microscope analysis ( Figure 3D), AnnexinV was detected at the cell membrane of fixed myoFB counterstained with Propidium Iodide (PI). To better discriminate between early and late apoptotic versus necrotic cells, cytograms of FC-AnnexinV versus PI were analysed. As shown in Figure 3E-F, untreated myoFB were mainly AnnexinV negative/ PI negative , indicating high viability and low apoptosis/necrosis. After acute treatment with increasing doses of NGF ( Figure 3E), a significant number of AnnexinV positive/ PI negative cells (early stages of apoptosis) with a low proportion of AnnexinV positive/ PI positive (late stages of apoptosis) was found. This increase was associated with NGF treatment. Upon chronic exposure ( Figure 3F), cells progressed to late stages of apoptosis, as detected by a significant increase of AnnexinV positive/ PI positive rate (Figure3C-D). The presence of AnnexinV positive/ PI negative cells in untreated myoFB, an indicator of early apoptotic myoFB, might be explained as a physiological event occurring in cultured myoFB.

p75 NTR -bearing MyoFB Undergo Apoptosis via Caspase3 Activation
As observed by studies of double staining, p75 NTR expressing myoFB showed round condensed chromatin (picnotic) nuclei. This TUNEL/p75 NTR coexpression was particularly evident for 100 ng/mL NGF treated myoFB, in comparison to TGFb1 treated ones ( Figure 4A). By itself, TGFb1 treatment did not significantly influence the number of apoptotic cells.
In these p75 NTR bearing cells, the contribution of caspase pathway was also investigated by using terminal effector caspase3 Figure 1. Characterization of TGFb1-induced myoFB. Confluent myoFB were exposed to NGF at the indicated doses/times and evaluated for biochemical and molecular changes. (A) aSMA protein expression in both myoFB and FB, as detected by cell surface ELISA (optic density) and flow cytometry (fluorescence intensity). (B-C) Fluorescent histograms showing a different pattern of trkA NGFR and p75 NTR expression in acute and chronic 100 ng/mL NGF exposed myoFB, as compared to untreated myoFB (black line). Mean Fluorescent Intensity (MFI) plots are shown inside fluorescent histograms (iso stands for isotype-matched control FI; 7.05 for trkA NGFR and 8.39 for p75 NTR ). (D) Morphological distribution of trkA NGFR /p75 NTR in untreated, acute and chronic 100 ng/mL NGF treated myoFB. (E-F) Target gene expression specific for trkA NGFR /p75 NTR after acute and for aSMA/ p75 NTR after chronic exposure to increasing NGF (p,.05). doi:10.1371/journal.pone.0047316.g001 antibodies. Overlays and single staining specific for cleaved (active) caspase3 or p75 NTR expression are shown in untreated, acute and chronic NGF exposed myoFB ( Figure 4B). The round morphology and the p75 NTR /cleaved caspase3 coexpression are particularly visible in chronic NGF exposed myoFB with respect to untreated myoFB. Flow cytometry showed that cleaved caspase3 was increased in NGF exposed myoFB, in a dose-dependent fashion, with a maximum effect at 100 ng/mL NGF. A selected p75 NTRbearing population exposed to 100 ng/mL NGF for 12 hrs and showing an increase of cleaved caspase3 is plotted in Figure 4C.
Since myoFB express both trkA NGFR /p75 NTR on their surface, NGF-induced apoptosis was also investigated in the presence of trkA NGFR and p75 NTR specific inhibitors, alone or in combination. As noted, active caspase3 signal was significantly reduced after preincubation with trkA NGFR and/or p75 NTR chemical inhibitors (500 ng/mL; Figure 4D). This effect suggests that both low and high trkA NGFR /p75 NTR rates are involved in the apoptotic process. A low degree of apoptotic signal was detected in both untreated and TGFb1-treated myoFB.

p75 NTR Inhibits and trkA NGFR Contributes to NGF-induced MyoFB Apoptosis
To provide further information on p75 NTR and particularly trkA NGFR /p75 NTR contribution in NGF-mediated apoptosis, appropriate silencing (siRNA) experiments were designed (see flow chart in Figure 5A). In preliminary studies, a dose-dependent increase of apoptotic signal (caspACE VAD) was detected after exposure to increasing NGF, resembling those of both AnnexinV and cleaved caspase3. Electroporation efficiency was estimated to be around 70% to 88% (counting GFP positive cells; Figure 5B). In accordance with the molecular data, flow cytometry analysis showed that a proportion of p75 NTR protein was still present in the culture system, as shown in Figure 5C-D. Confluent control-siRNA and p75 NTR -siRNA electroporated cells were exposed to 100 ng/mL NGF or 10 ng/mL TGFb1 and harvested after 3 days to evaluate and compare the apoptotic signals. As shown in Figure 5E, electroporation with 1.5 mg/mL p75 NTR -siRNA was effective in reducing the apoptotic signal (35%-decrease in apoptosis) in 100 ng/mL NGF treated cells, as compared to 100 ng/mL NGF-treated control siRNA cells (referred as 100% apoptosis). Control-siRNA had no effect on apoptotic signal. This low but significant reduction of apoptotic signal might be explained by the presence of trkA NGFR . Therefore to verify this explanation, a double silencing of both trkA NGFR and p75 NTR was performed next and electroporated cells were exposed to 100 ng/ mL NGF or 10 ng/mL TGFb1 (positive control). Even in the presence of both receptor silencings, apoptosis was not completely abolished in 100 ng/mL NGF treated cells (47-50%-decrease in apoptosis over untreated cells (100%)). The results demonstrate that p75 NTR -siRNA is effective in reducing the apoptotic signal induced in NGF-treated myoFB, while trkA NGFR -silencing might make some contribution. A slight apoptotic signal was also detected in untreated myoFB and TGFb1-treated ones.

Discussion
The main finding of this study indicates that TGFb1-induced conjunctival myoFB (herein referred as myoFBs) significantly increase their p75 NTR and undergo apoptosis upon acute and particularly chronic NGF treatment. Both chemical inhibition and single-(p75 NTR ) and double (trkA NGFR /p75 NTR )-silencing approaches (siRNA) down-regulated the apoptotic signal, highlighting the contribution of NGF/p75 NTR in mediating myoFB apoptosis.
Tissue remodeling and fibrosis clearly compromise ocular surface structure and lead to ocular function decline [1,2]. FB/ myoFB, known to drive healing response, can be modulated at various levels. The injured ocular microenvironment as well as tears and aqueous humor contain several cytokines, growth factors and chemical mediators that greatly power the local response, contributing as profibrogenic mediators [2]. We previously reported the NGF ability to stimulate in vitro the induction of myoFB phenotype and matrix contraction, highlighting NGF contribution at both early and late stages of proper (physiological) resolution of tissue repair [13,14,[18][19][20][21]. Focused studies on these FB demonstrated that all the profibrogenic NGF effects were partially mediated by modulation of TGFb1 [13,14,18], in line with findings provided in other systems as well as in animal models [22,23].
In order to investigate the NGF effect on myoFB, a cell culture model of TGFb1-induced myoFB was reproduced and firstly characterized for NGF receptor pathway. TGFb1 is a widely accepted chief inducer of FB differentiation, even if overlapping mechanisms and soluble factors might contribute massively to the development of myoFB phenotype in inflamed tissues, [17]. In this model, FB differentiate into myoFB upon TGFb1 exposure, shifting their morphology from a typical flattened irregular shape (FB) to an elongated and spindle shaped appearance (myoFB), associated with a consistent aSMA expression [3]. These myoFB showed a significant increase of aSMA and p75 NTR expression, and preserved trkA NGFR expression, in comparison to quiescent aSMA negative conjunctival FB expressing only trkA NGFR and/or low levels of p75 NTR [11,24]. To the best of our knowledge, this is the first evidence of TGFb1 as an inducer of trkA NGFR /p75 NTR expression, while NGF induction of TGFb1 has been reported for other cell types [13,14]. p75 NTR expressing myoFB have been detected in fibrotic tissues, where myoFB differentiation is mainly mediated by TGFb1 [18,25].
Next, the NGF receptor pathway was investigated in stimulating and neutralizing experiments. Acute and massively chronic NGF exposure triggered a selective increase of p75 NTR expression while trkA NGFR expression was slightly modulated. Both treatments shifted the trkA NGFR /p75 NTR rate toward p75 NTR expression, in a dose-and time-dependent manner, resulting in a decrease of the trkA NGFR /p75 NTR ratio. As previously reported, p75 NTR expression is a typical feature of healing myoFB [20,24]. Indeed, NGF effect was associated with a significant decrease of aSMA and TGFb1 profibrotic genes, the main markers of myoFB survival. An inverse correlation was detected between aSMA and p75 NTR gene expression. The higher trkA NGFR /p75 NTR ratio in untouched FB (data not shown), the slight decrease in induced myoFB and the significant decreased upon NGF exposure in favor of p75 NTR expression, suggest a possible NGF-p75 NTR involvement in myoFB survival.
As previously reported, conditions leading to a lower trkA NGFR / p75 NTR ratio are often associated with apoptosis [12,15,24]. An increase in the number of rounded/condensed cells with shrinkage of cytoplasm (monolayers) and several free-floating cells (supernatant) were observed after acute and chronic NGF treatments. As confirmed by in situ end labeling of oligonucleotyde fragments (TUNEL), a proportion of myoFB had fragmented DNA, an effect not observed when cells were exposed to TGFb1 [26]. To discriminate early and late apoptotic cells from necrotic ones, a specific AnnexinV/PI analysis was carried out: early apoptotic cells were selected from live and dead cells by cell sorting of AnnexinV labeling and exclusion of vital dye. This AnnexinV/PI study extended the above mentioned results showing an increased number of late apoptotic cells upon chronic treatment. Since NGF effects were studied in myoFB cultured in medium supplemented with 1% FCS, mimicking the microenvironment rich in growth factors, a counteracting serum effect on NGF apoptosis should not be excluded. More interestingly, AnnexinV positivity was drastically reduced in cells treated in the presence of neutralizing anti-NGF antibodies or in cells pretreated with chemical receptor neutralizers, suggesting a task for NGF and its receptors at the resolution of the repair process.
Arising from quiescent FB and transiently expressed during wound-healing, myoFB play a crucial role in the late phases of the repair process being responsible for the ECM deposition and contraction [2][3][4]17]. MyoFB disappearance has been documented during remodeling stages or in fibrotic tissues, suggesting that a gene-directed cellular self-destruction (apoptosis) is necessary for resolution of fibrosis [27]. Unwanted persistence of myoFB could be one of the causes for excessive scarring and fibrosis [28,29]. MyoFB resistance to apoptosis seems to be strongly regulated by cytokines and growth factors, which might influence the microenvironment [5]. TGFb1 is unequivocally recognized as inducing survival of myoFB both in vitro and in vivo, even though other growth factors and cytokines might contribute significantly [5,26]. The observation that apoptosis paralleled TGFb1 and aSMA down-regulations might suggest a possible NGF modulation of myoFB survival directly by inducing apoptosis or indirectly by decreasing TGFb1 levels [29]. In line with this observation, TGFb1 is also a survival factor for myoFB and is always increased in fibrotic tissues [5,17,26]. In the context of in vivo fibrotic-tissue system, we might hypothesize that NGF might trigger ''selective'' apoptosis, after being released by (primed)-FB/myoFB and other similar cell types populating the wounded area. From our study, TGFb1 continued to stimulate NGF productions (data not shown). This hypothesis is in line with previous studies indicating that fetal FB express high amounts of NGF and scarring phenomena not present during the fetal period and that NGF topical application allows proper healing and does not trigger fibrosis in injured tissues [30,9,19,23].
In our study, the cells showing DNA fragmentation were also recognized as p75 NTR bearing cells, as detected by sorting analysis. NGF exposure resulted in increasing cleavage of caspase-3 significantly expressed in p75 NTR bearing myoFB. Even though in this cell culture model, the apoptotic pathway induced by NGF appears associated with activation of the effector caspase-3, the intracellular death-pathway induced by p75 NTR activation is unknown.
According to the literature, trkA NGFR mainly drives proliferation, differentiation and survival while p75 NTR mainly triggers differentiation and apoptosis [12,15]. Specific single/ double neutralizing experiments showed the contribution of both receptors in PS outer-expression and caspase-activation. As an explanation, both receptors are present on the surface either in homo and/or hetero dimeric state and depending on the magnitude/specificity of the binding, both receptors can drive a response [15,16].
Therefore to provide additional information, apoptosis was also investigated in NGF-treated myoFB expressing low trkA NGFR / p75 NTR ratio (100 ng/mL) by using small inhibitory RNAs (siRNA) directed against p75 NTR [31]. Specific silencing of p75 NTR significantly reduced the percentage of apoptosis after 100 ng/mL NGF exposure, as observed in comparison to control siRNA myoFB. However, a complete reduction of apoptotic signal was not detected. According to the literature, p75 NTR seems to allow apoptosis in receptive cells either directly (ceramide and caspase activation) or indirectly by the modulation of trkA NGFR trafficking/signaling [15,16,32]. Indeed, trkA NGFR -mediated apoptosis has been also documented, ascribed to a novel Ras and/or Raf signaling pathway [33]. Since trkA NGFR is primarily, but not exclusively, a proliferating and survival receptor, cells were subjected to double knocking-down experiments. Still in the presence of double trkA NGFR /p75 NTR silencing, apoptosis was as low as those of single p75 NTR silencing, but not completely abolished. In our studies of silencing, both single and double trkA NGFR /p75 NTR target gene amplification confirmed only a partial decrease in mRNA expression. Besides fitting transfection efficiency, some proteins still persist in the system even after 72 hrs from gene knocking down: a change in the RNA level for a particular gene product does not directly correlate with a change in the amount of protein in the cell. This explanation could fit our results to explain the incomplete p75 NTR and trkA NGFR /p75 NTR protein silencing. An additional explanation is that in this system, NGF might not be the only factor responsible for myoFB induced apoptosis. Taken together, the explanations for a 37%-decrease in apoptosis during single-and 47%-decrease in apoptosis during double-silencing RNA experiments might be due to an incomplete silencing of technique, an intracellular p75 NTR retaining, an enhanced trkA NGFR and/or p75 NTR signaling, or other indirect factors induced by NGF signaling [8,16,23,32,33]. Several data on experimental models showed the selective expression of p75 NTR in myoFB within fibrotic tissues and the p75 NTR contribution to myoFB differentiation and cell apoptosis [25,[34][35][36][37]. At least in part, these in vitro results might provide an explanation for the observation that NGF does not induce a massive apoptotic effect on fibrotic tissues. Pictures depicting TUNEL reactivity in p75 NTR positive cells exposed to chronic 100 ng/mL NGF, as compared to TGFb1 exposed cells. Changes in morphology, such as having rounded cytoplasm and markedly condensed chromatin/nuclear fragmentation, are clear visible in p75 NTR -bearing myoFB. A few TUNEL-positive myoFB were also quantified in untreated cultures (not shown). (B) Overlays and single staining specific for p75 NTR and cleaved caspase3 are visible in both acute and chronic 100 ng/mL NGF exposed myoFB, as compared to untreated myoFB. (C-D) p75 NTR positive sorted cells analysed for active caspase3 expression and related statistical analysis of MFI. Note the specific reduction upon pretreated with trkA NGFR and/or p75 NTR chemical inhibitors. doi:10.1371/journal.pone.0047316.g004 Taken together, our in vitro model provides data on NGF as an inducer of apoptosis by shifting the trkA NGFR /p75 NTR ratio in favor of p75 NTR and inducing up-regulation of terminal caspase-3, opening a more fruitful possibility for counteracting myoFB resistance to apoptosis. Indeed, the observed NGF-mediated TGFb1 deregulation is in line with the apoptotic effect and with the inhibition of TGFb1-mediated myoFB endurance. In line, our data might explain the observation of the successful corneal healing upon NGF topical application [9,19]. It still remains to be elucidated whether during chronic inflammatory responses characterized by increased p75 NTR and NGF levels, NGF might improve apoptosis of myoFB in the presence of pro-and/or antiapoptotic factors (TGFb1, IL1b, PDGF, FGF, IL4, IL6) [38][39]. Cumulatively, these findings would suggest new therapeutic strategies to offset overt fibrosis by modulating myoFB survival in fibrotic tissues.

The myoFB Model
The research protocol followed the tenets of the Declaration of Helsinki and was approved by the Intramural Committee of Hadassah University Hospital. Human primary cultures of conjunctival FB (passages 3-7) were obtained from Innoprot (Bizkaia, Spain) and used for these studies. Cell cultures were expanded (24 hrs doubling time) in DMEM supplemented with 10% heat-inactivated Fetal Calf Serum (FCS), 2 mM-glutamine, 100 U/ml pen-streptomycin, under standard culturing conditions and re-plated after trypsin-EDTA harvesting. To obtain in vitro myoFB, preliminary titration experiments were carried out with increasing TGFb1 concentrations (0.1, 1, 2, 5, 10 and 20 ng/mL) in serum free conditions [40]. Both molecular and biochemical analysis of aSMA and myoFB/FB rate expression identified 2 ng/ mL TGFb1 over 3d for almost 70-80% myoFB transdifferentiation. In some experiments, enriched myoFB (99% myoFB) were obtained after selection (EasySepHMagnet, Stem Cell Technology, Voden Medical Instruments, Milan, Italy) with anti-aSMA antibodies conjugated to magnetic biodegradable beads (Miltenyi Biotec, GmbH, Germany), according to the manufacturer's procedure.

Experimental Procedure
TGFb1-induced myoFB (herein termed myoFB and defined as aSMA-expressing myoFB) [17,40], were re-plated at high density in the absence of FCS to retain aSMA phenotype. At confluence, monolayers were exposed to increasing doses of NGF in the presence of 1% FCS (to guarantee synchronization and reduction of autophosphorylation) over the specified time-points. Acute and chronic NGF exposures were performed as follows: acute exposure was a single stimulus and chronic was three stimuli repeated every two days (NGF replaced in fresh medium). Acute and chronic samplings were performed at 5, 12, 24 and 48 hrs from last stimulus, depending on the evaluation. For neutralizing experiments, monolayers were exposed to chemical inhibitors or specific antibodies for 30 min prior to the addition of NGF. The startingtime point (control) indicates untreated myoFB re-plated and harvested in parallel with NGF treated myoFB. Cells were trypsin harvested and washed in Hank's Balanced Salt Solution (HBSS without Ca 2+ /Mg 2+ ), before biochemical or molecular evaluations. For most of the experiments, TGFb1 exposure was carried out in parallel, according to the survival effects of TGFb1 on myoFB [26].

MTS Assay and Cell-surface ELISA
Cells were seeded on 96-well plates and then treated with increasing NGF doses and times depending on the experiment. Cell viability was determined by measuring mitochondrial reduction of the MTS dye reagent into a soluble formazan product, according to the manufacturer's instructions (Promega, Madison, WI). Cell surface ELISA was carried out for quantifying aSMA expression, as previously reported in detail [13,14,18]. Absorbance measurements (490 nm) were recorded using a Sunrise spectrophotometer (Tecan Systems, Inc., San Jose, CA).

Confocal Laser Microscopy
Cells on round coverslips (2x10 5 ) were washed in 10 mM Phosphate-Buffered Saline (PBS, pH7.5) at the stated time-points, fixed in 3.7% buffered PFA, quenched in 50 mM NH 4 Cl to rule out autofluorescence and blocked/permeabilized in 3% BSA and 0.03% Triton X-100 (TX) in PBS. Monolayers were probed with: rabbit anti-trkA NGFR antibodies and goat anti-p75 NTR antibodies (4 mg/mL, both from R&D); mouse anti-aSMA (2 mg/mL; Santa Cruz) and active caspase3 (Cell signaling). Specific binding of the primary antibody was detected using Cy2 or APC-conjugated secondary antibodies (1/500). Nuclear counter-staining was performed with Propidium Iodide (25 mg/mL) or with TOTO-3 (1 mM) in PBS containing RNase (20 mg/mL) and coverslips were mounted in hand-made anti-fade medium. Irrelevant isotypematched IgG antibodies (Vector) were incubated in parallel and used as internal controls for the channel series acquisitions and related background subtraction. Monolayers were examined and images were acquired using an inverted E2000-U microscope equipped with the C1 software (x20/0.45NA; x40/0.60NA; x60/ 1.4 oil; Nikon, Tokyo, Japan). TIFF-converted pictures were assembled by Adobe Photoshop 7.0 (Adobe Systems Inc., San Jose, CA).

Relative Real-Time PCR
Total RNA was extracted from cells (2610 5 ) treated and harvested at 5 or 24 hrs from last stimulation, using the EuroGold TRIfast TM kit (Euroclone, Milan, Italy) and diluted in RNase free water (Millipore). The concentration and purity (260/280 nm and 260/230 nm) of total RNA were determined by using a spectrophotometer (NanoDropH ND-1000, Wilmington, DE, USA), while 28S/18S ratios were recorded after agarose gel separation. When required, a DNAse treatment was allowed to guarantee absence of contaminating DNA (Turbo DNA free kit; Ambion, Milan, Italy). Only RNAs showing RNA/DNA rate .1.8 were used for analysis. Real time PCR was performed in a two step manner: cDNA synthesis and amplification were carried out in a One Personal thermocycler (PeqLab, EuroClone) and in an Opticon2 system (MJ Research, Watertown, MA), respectively. Oligo dT 21 -primer was used to generate cDNA from 1 mg total RNA, according to the manufacturer's instructions (Improm kit; Promega). SYBR-Green HotStart AmpliGold Taq polymerase (Applied Biosystems, Foster City, CA) was used for specific amplification of 3 mL cDNA in a final volume of 20 mL in 96-well plate including internal controls, at stated amplification settings (see Table 1). In preliminary studies, electrophoresed amplicons were purified with a gel extraction kit (Wizard SV gel system, Promega) and sequenced on a DNA sequencing system (ABI PRISM 3700; Applied Biosystems). Each RT-PCR reaction contained equivalent amounts of total RNA (normalized samples). Target gene expression ratio was calculated from all single Cts (both target and referring genes in duplicate per sample) run in the Relative Expression Software Tool REST-384 ß ver.2 [41]. Data are expression ratios in 2log scale relative to treated vs. untreated myoFB and untreated myoFB vs. untreated FB. As negative controls, amplification was performed in the presence of total RNA (exclusion of genomic contamination) or on cDNAs obtained from reverse transcription tubes in the absence of enzyme (exclusion of unspecific amplification).

Apoptosis Detection
Conditioned media were gently removed and centrifuged to collect death cells, and adherent cells were briefly trypsin harvested and processed as reported in detail.
Hoechst/DAPI staining. Cells (2x10 5 ) were treated with NGF for 5 hrs and stained with 1 mg/mL Hoechst 33342 (HO342) for 5 min at 37uC or 1 mg/mL DAPI for 15 min at 37uC (both from Invitrogen-Molecular Probes, Milan, Italy) to identify cells with membrane shrinkage, chromatin condensation and clumping or enlarged cells. Monolayers were photographed with a microscope equipped with epifluorescence to detect Hoechst/DAPI signals and software for image acquisition (Nikon, Tokyo, Japan). TIFF-converted pictures were assembled by Adobe Photoshop 7.0.
TUNEL staining. Adherent cells (2610 5 ) were stained with the biotinyled dATP nick end labeling procedure (TUNEL), which mark DNA fragmentation. Briefly, fixed (3.7% PFA) and permeabilized (0.1% TX) cells were incubated with exogenous rTdT enzyme and biotin 14-dATP (37uC/3 hrs; Invitrogen), for repair of 39-hydroxyl DNA ends. Positive controls were carried out in parallel, with a DNase I pre-incubation of monolayers (2 U/ mL; 37uC/30 min). Negative controls were monolayers that were incubated with buffer lacking rTdT enzyme. The apoptotic nuclei were labeled according to the ABC technique (Vector Laboratories, Burlingame, CA) using DAB solution (Dako Corp., Carpinteria, CA) as substrate. TUNEL positive cells were scored by two independent observers (blind fashion), and five optic fields/ coverslip for at least 10 coverslips were counted for each culture condition (x20). The percentage of apoptosis was calculated as (number of apoptotic cells/number of total cells)6100. For double staining with p75 NTR and specific FC-secondary antibodies, TUNEL reaction was performed according to the TRITC-Avidin procedure (Vector). Stained cells were imaged as reported in the confocal laser microscopy section.
FACS analysis of apoptosis. Single harvested cells (10 6 ) were equilibrated in Hepes buffer containing 2 mM CaCl 2 , and subjected to immunofluorescence as follows. To detect externalized PS, cells were probed with FC-AnnexinV-Apoptosis detection kit (Miltenyi), according to the manufacturer's directions with the exception that cells were post treated with RNAse A and finally fixed in 3% PFA [42,43]. Controls were prepared according to the recommendation of the supplier. Single harvested cells were double-stained with cleaved caspase3 and APC conjugated anti-Rabbit secondary antibodies and FC-conjugated anti-p75 NTR antibody in FACS buffer. Logical gate (back-gating on FL2-FC) combining p75 NTR positive cells and their scatter properties were used for select cleaved caspase-3 positive cell population. Ca-spACE-VAD and cleaved caspace3 quantifications were carried out according to the manufacturer's instructions (Promega). Acquisitions and analyses were carried out as above described in detail in the FACS session.

Small Interference RNA (siRNA)
Specific Stealth TM siRNA molecules were chemically synthesized, as desalted, 25mer duplex oligonucleotides (Table 1) (rnaidesigner.invitrogen.com/). The Nucleofection was performed in an electroporation machine (Amaxa GmbH; Instrumentation Laboratory, Milan, Italy). To set up the technique, preliminary experiments were carried out according to the optimized protocols of the manufacturer's instructions (Amaxa's silencing kit). Three different silencing concentrations (0.5, 1 and 1.5 mg/mL) were tested and 5-72 hrs after plating, electroporated cells were screened for the highest % of efficiency (80-90%) in knock-down gene expression, by specific mRNA (PCR) and protein (FACS) analysis. For serial specific experiments (1.5 m/mL siRNA): singlecells (10 6 ) were re-suspended in 100 mL specific buffer (under serum-free conditions), mixed with GFP marker and siRNA sequences [single (p75 NTR -siRNA) or both (trkA NGFR /p75 NTR -siRNA) or non-specific (control siRNA)], quickly nucleofected, harvested in 1 mL pre-warmed culturing medium and seeded in 6 well plates or T21cm2 plates or 4-chamber slides (Nunc) depending on analysis. Transfected cells were cultured and fed daily with fresh medium until they were stimulated and assayed (flow-chart in Figure 5A).

Statistical Analysis
All the experiments were done independently at least three times, each experiment being carried out in duplicate or triplicate. Data (means6SD) were exported to Excel datasheet and presented as graphs (Excel Microsoft Corp., Redmond, WA; Microsoft.com/excel). Statistical analyses were performed by using the StatView II package for PC (Abacus Concepts. Inc., Berkeley, CA). All data followed a normal distribution and parametric ANOVA followed by Tukey-Kramer post hoc was employed to analyze data. The Spearman's rank correlation coefficient (Rho) was calculated to identify specific target relationships. A confidence of 95% was presumed to reflect significant difference between group mean values, and therefore a p-value of ,.05 was taken as the limit of significance. A specific REST/ANOVA coupled analysis was carried out for PCR experiments.